Neural Networks for Machine Learning Lecture 13a The ups ...
嚜燒eural Networks for Machine Learning
Lecture 13a
The ups and downs of backpropagation
Geoffrey Hinton
Nitish Srivastava,
Kevin Swersky
Tijmen Tieleman
Abdel-rahman Mohamed
A brief history of backpropagation
?
The backpropagation algorithm
for learning multiple layers of
features was invented several
times in the 70*s and 80*s:
每 Bryson & Ho (1969) linear
每 Werbos (1974)
每 Rumelhart et. al. in 1981
每 Parker (1985)
每 LeCun (1985)
每 Rumelhart et. al. (1985)
?
?
Backpropagation clearly had great
promise for learning multiple layers
of non-linear feature detectors.
But by the late 1990*s most serious
researchers in machine learning had
given up on it.
每 It was still widely used in
psychological models and in
practical applications such as
credit card fraud detection.
Why backpropagation failed
?
The popular explanation of why
backpropagation failed in the 90*s:
每 It could not make good use of
multiple hidden layers.
?
(except in convolutional nets)
每 It did not work well in recurrent
networks or deep auto-encoders.
每 Support Vector Machines worked
better, required less expertise,
produced repeatable results,
and had much fancier theory.
?
The real reasons it failed:
每 Computers were thousands
of times too slow.
每 Labeled datasets were
hundreds of times too small.
每 Deep networks were too small
and not initialized sensibly.
These issues prevented it from
being successful for tasks where
it would eventually be a big win.
A spectrum of machine learning tasks
Typical Statistics---------------Artificial Intelligence
?
?
?
?
Low-dimensional data
(e.g. less than 100 dimensions)
Lots of noise in the data.
Not much structure in the data.
The structure can be captured
by a fairly simple model.
The main problem is separating
true structure from noise.
每 Not ideal for non-Bayesian
neural nets. Try SVM or GP.
?
?
?
?
High-dimensional data (e.g. more
than 100 dimensions)
The noise is not the main problem.
There is a huge amount of structure
in the data, but its too complicated to
be represented by a simple model.
The main problem is figuring out a
way to represent the complicated
structure so that it can be learned.
每 Let backpropagation figure it out.
Why Support Vector Machines were never a good bet for
Artificial Intelligence tasks that need good representations
?
View 1: SVM*s are just a clever ? View 2: SVM*s are just a clever
reincarnation of Perceptrons.
reincarnation of Perceptrons.
每 They expand the input to a
每 They use each input vector in
(very large) layer of nonthe training set to define a
linear non-adaptive features.
non-adaptive ※pheature§.
? The global match between a
每 They only have one layer of
test input and that training input.
adaptive weights.
每 They have a clever way of
每 They have a very efficient
simultaneously doing feature
way of fitting the weights
selection and finding weights on
that controls overfitting.
the remaining features.
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related searches
- probability for machine learning pdf
- machine learning neural networks
- machine learning vs neural network
- neural networks for dummies
- artificial neural networks background
- neural networks ai
- neural networks from scratch pdf
- types of neural networks pdf
- graph neural networks ppt
- artificial neural networks pdf free
- neural networks and learning machines
- learning convolutional neural networks for graphs